No abstract

The present study is the second in a series aiming at a systematic inventory of specific toxic effects of oils. By employing a recombinant yeast stably transfected with human estrogen receptor-alpha (ERalpha) or -beta (ERbeta) or androgen receptor (AR) and expressing yeast enhanced green fluorescent protein, the (anti-)estrogenicity and (anti-)androgenicity of 11 crude oils and refined products were studied. None of the oils tested had significant estrogenic effects in the ERalpha assay or androgenic effects in the AR assay. However, all oils were capable of inducing estrogenic responses in the ERbeta assay, with several responses being above even the maximal response of the standard 17beta-estradiol (E2). Based on the lowest effect concentrations, the potencies of oils in all the assays were between four and seven orders of magnitude lower than those of the standards E2 or testosterone (T). The potencies of the actual individual petrochemical agonists may, however, be relatively high, considering the complex composition of oils. Additive effects, antagonistic effects, and a synergistic effect were measured in the assays upon coexposure to a fixed concentration of standard (E2 or T) and increasing concentrations of oils. To investigate whether the observed effects were receptor-mediated, coexposures to the synthetic inhibitors ICI 182,780 (ERbeta assay) or flutamide (AR assay), a fixed concentration of standard, and various concentrations of oils were performed. The results suggested that the androgenic effects were receptor mediated, whereas the estrogenic effects may be only partially mediated via the receptor. The present study indicates that oils contain compounds with possible endocrine-disrupting potential, some of them acting via the hormone receptors.

The present study is the first in a series reporting on in vitro toxic potencies of oils. The objective was to determine whether 11 crude oils and refined products activate the aryl hydrocarbon receptor (AhR) in a dioxin receptor-mediated luciferase assay. Cells were exposed for 6 and 24 h to different oil concentrations to screen for polycyclic aromatic hydrocarbon-like or dioxin-like activity. Moreover, cytotoxicity of the oils was determined using rat hepatoma cells. Except for one crude oil, none of the oils appeared cytotoxic up to 100 mg/L, but all oils activated the AhR. Strong AhR induction was observed for most oils after 6 h, and responses decreased after 24 h, indicating the presence of metabolizable agonists. However, several oils still caused high responses after 24 h, also demonstrating the presence of persistent agonists. The potencies (calculated based on comparisons of concentrations at which 50% of the maximal effect was observed) of oils were found to be approximately 40 to 106 times lower than the potency of the assay's standards benzo[a]pyrene and 2,3,7,8-tetrachlorodibenzo-p-dioxin. However, considering that oils contain thousands of chemicals, the potencies of petrochemical agonists may be very high. Among the most potent oils were bunker and crude oils. Induction up to 200% as compared to the maximum induction caused by benzo[a]pyrene was observed for these oils. Such supermaximal responses suggest mixture effects that may not be receptor-mediated. Experiments in which oils were tested in combination with the standards demonstrated that oils acted via an antagonistic or additive mode. The results of the present study may help improve risk assessment of petroleum products and judge the necessity or priority of oil spill cleanup activities.

Recent studies have indicated that in addition to narcosis certain chemicals in crude oils and refined petroleum products may induce specific modes of action, such as aryl hydrocarbon receptor (AhR) agonism. The risks these toxic compounds pose to organisms depend on internal exposure levels, as driven by the chemicals' bioaccumulation potential. Information on this potential however is lacking, as the chemicals' identity mostly is unknown. This study showed that AhR agonists bioaccumulate from oil-spiked sediments into aquatic worms and persist in the worms for at least several weeks. Chemical fractionations of eight pure oils into saturates, aromatics, resins, and asphaltenes (SARA), followed by effect-directed analyses using in vitro reporter gene assays revealed that the agonists predominantly are aromatic and resin-like chemicals. Some of the compounds were easily metabolized in vitro, while others were resistant to biotransformation. HPLC-assisted hydrophobicity profiling subsequently indicated that the AhR-active chemicals had a high to extremely high bioaccumulation potential, considering their estimated logK(ow) values of 4 to >10. Most of the AhR agonism, however, was assigned to compounds with logK(ow) of 5-8. These compounds were present mainly in the mid to high boiling point fractions of the oils (C(14)-C(32) alkane range), which are usually not being considered (the most) toxic in current risk assessment. The fractionations further revealed considerable oil and fraction-dependent antagonism in pure oils and SARA fractions. The results of this study clearly demonstrate that crude oils and refined petroleum products contain numerous compounds that can activate the AhR and which because of their likely persistence and extremely high bioaccumulation potential could be potential PBT (persistent, bioaccumulative and toxic) or vPvB (very persistent and very bioaccumulative) substance candidates. Many chemicals were identified by GC-MS, but the responsible individual compounds could not be exactly identified in the complex mixtures of thousands of compounds. Because this obstructs a classical PBT risk assessment, our results advocate an adapted risk assessment approach for complex mixtures in which low concentrations of very potent compounds are responsible for mixture effects.

Current petroleum risk assessment considers only narcosis as the mode of action, but several studies have demonstrated that oils contain compounds with dioxin-like, estrogenic or antiestrogenic, and androgenic or antiandrogenic activities. The present study is the third in a series investigating the specific toxic effects of 11 crude oils and refined products. By employing recombinant mammalian cells stably transfected with the human estrogen receptor alpha (ERα) or beta (ERβ), and expressing the luciferase protein (ERα-U2OS-Luc and ERβ-U2OS-Luc assay), the estrogenicity or antiestrogenicity of oils was studied. All oils, except for two refined oils and one crude oil, induced estrogenic responses. The calculated estrogenic potencies of the oils were six to nine orders of magnitude lower than the potency of 17β-estradiol (E2). Upon coexposure to a fixed concentration of E2 and increasing concentrations of oils, additive, antagonistic, and synergistic effects were revealed. One nautical fuel oil was tested in the human breast carcinoma cell line MCF-7, in which it induced cell proliferation up to 70% relative to the maximal induction by E2. At its minimum effect concentration of 25 mg/L, the oil was also capable of inducing mRNA expression of the estrogen-dependent protein pS2 by a factor of two. The present results indicate that oils naturally contain potentially endocrine-disrupting compounds that are able to influence the estrogenicity of other compounds and may cause biological responses beyond receptor binding.